openaivec 0.15.1__tar.gz → 0.99.0__tar.gz

{openaivec-0.15.1 → openaivec-0.99.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: openaivec
-Version: 0.15.1
+Version: 0.99.0
 Summary: Generative mutation for tabular calculation
 Project-URL: Homepage, https://microsoft.github.io/openaivec/
 Project-URL: Repository, https://github.com/microsoft/openaivec
@@ -32,6 +32,27 @@ Description-Content-Type: text/markdown
 
 **openaivec** enables data analysts to seamlessly integrate OpenAI's language models into their pandas and Spark workflows. Process thousands of text records with natural language instructions, turning unstructured data into actionable insights with just a few lines of code.
 
+## Contents
+- [Why openaivec?](#why-openaivec)
+- [Quick Start](#-quick-start-from-text-to-insights-in-seconds)
+- [Real-World Impact](#-real-world-impact)
+- [Overview](#overview)
+- [Core Workflows](#core-workflows)
+- [Using with Apache Spark UDFs](#using-with-apache-spark-udfs)
+- [Building Prompts](#building-prompts)
+- [Using with Microsoft Fabric](#using-with-microsoft-fabric)
+- [Contributing](#contributing)
+- [Additional Resources](#additional-resources)
+- [Community](#community)
+
+## Why openaivec?
+- Drop-in `.ai` and `.aio` DataFrame accessors keep pandas analysts in their favorite tools.
+- Smart batching (`BatchingMapProxy`) deduplicates prompts, enforces ordered outputs, and shortens runtimes without manual tuning.
+- Built-in caches, retry logic, and reasoning model safeguards cut noisy boilerplate from production pipelines.
+- Ready-made Spark UDF builders and Microsoft Fabric guides take AI workloads from notebooks into enterprise-scale ETL.
+- Pre-configured task library and `FewShotPromptBuilder` ship curated prompts and structured outputs validated by Pydantic.
+- Supports OpenAI and Azure OpenAI clients interchangeably, including async workloads and embeddings.
+
 ## 🚀 Quick Start: From Text to Insights in Seconds
 
 Imagine analyzing 10,000 customer reviews. Instead of manual work, just write:
@@ -111,21 +132,28 @@ This approach helps reduce latency and simplifies your code.
 Additionally, it integrates effortlessly with Pandas DataFrames and Apache Spark UDFs, making it easy to incorporate
 into your data processing pipelines.
 
-## Features
+Behind the scenes, `BatchingMapProxy` and `AsyncBatchingMapProxy` deduplicate repeated inputs, guarantee response order,
+and unblock waiters even when upstream APIs error. Caches created via helpers such as `responses_with_cache` plug into
+this batching layer so expensive prompts are reused across pandas, Spark, and async flows. Progress bars surface
+automatically in notebook environments when `show_progress=True`.
+
+## Core Capabilities
 
-- Vectorized API requests for processing multiple inputs at once.
-- Seamless integration with Pandas DataFrames.
-- A UDF builder for Apache Spark.
-- Compatibility with multiple OpenAI clients, including Azure OpenAI.
+- Vectorized request batching with automatic deduplication, retries, and cache hooks for any OpenAI-compatible client.
+- pandas `.ai` and `.aio` accessors for synchronous and asynchronous DataFrame pipelines, including `ai.extract` helpers.
+- Task library with Pydantic-backed schemas for consistent structured outputs across pandas and Spark jobs.
+- Spark UDF builders (`responses_udf`, `embeddings_udf`, `parse_udf`, `task_udf`, etc.) for large-scale ETL and BI.
+- Embeddings, token counting, and similarity utilities for search and retrieval use cases.
+- Prompt tooling (`FewShotPromptBuilder`, `improve`) to craft and iterate production-ready instructions.
 
 ## Key Benefits
 
-- **🚀 Performance**: Vectorized processing handles thousands of records in minutes, not hours
-- **💰 Cost Efficiency**: Automatic deduplication significantly reduces API costs on typical datasets
-- **🔗 Integration**: Works within existing pandas/Spark workflows without architectural changes
-- **📈 Scalability**: Same API scales from exploratory analysis (100s of records) to production systems (millions of records)
-- **🎯 Pre-configured Tasks**: Ready-to-use task library with optimized prompts for common use cases
-- **🏢 Enterprise Ready**: Microsoft Fabric integration, Apache Spark UDFs, Azure OpenAI compatibility
+- **🚀 Throughput**: Smart batching and concurrency tuning process thousands of records in minutes, not hours.
+- **💰 Cost Efficiency**: Input deduplication and optional caches cut redundant token usage on real-world datasets.
+- **🛡️ Reliability**: Guardrails for reasoning models, informative errors, and automatic waiter release keep pipelines healthy.
+- **🔗 Integration**: pandas, Spark, async, and Fabric workflows share the same API surface—no bespoke adapters required.
+- **🎯 Consistency**: Pre-configured tasks and extractors deliver structured outputs validated with Pydantic models.
+- **🏢 Enterprise Ready**: Azure OpenAI parity, Microsoft Fabric walkthroughs, and Spark UDFs shorten the path to production.
 
 ## Requirements
 
@@ -145,7 +173,7 @@ If you want to uninstall the package, you can do so with:
 pip uninstall openaivec
 ```
 
-## Basic Usage
+## Core Workflows
 
 ### Direct API Usage
 

{openaivec-0.15.1 → openaivec-0.99.0}/README.md

@@ -6,6 +6,27 @@
 
 **openaivec** enables data analysts to seamlessly integrate OpenAI's language models into their pandas and Spark workflows. Process thousands of text records with natural language instructions, turning unstructured data into actionable insights with just a few lines of code.
 
+## Contents
+- [Why openaivec?](#why-openaivec)
+- [Quick Start](#-quick-start-from-text-to-insights-in-seconds)
+- [Real-World Impact](#-real-world-impact)
+- [Overview](#overview)
+- [Core Workflows](#core-workflows)
+- [Using with Apache Spark UDFs](#using-with-apache-spark-udfs)
+- [Building Prompts](#building-prompts)
+- [Using with Microsoft Fabric](#using-with-microsoft-fabric)
+- [Contributing](#contributing)
+- [Additional Resources](#additional-resources)
+- [Community](#community)
+
+## Why openaivec?
+- Drop-in `.ai` and `.aio` DataFrame accessors keep pandas analysts in their favorite tools.
+- Smart batching (`BatchingMapProxy`) deduplicates prompts, enforces ordered outputs, and shortens runtimes without manual tuning.
+- Built-in caches, retry logic, and reasoning model safeguards cut noisy boilerplate from production pipelines.
+- Ready-made Spark UDF builders and Microsoft Fabric guides take AI workloads from notebooks into enterprise-scale ETL.
+- Pre-configured task library and `FewShotPromptBuilder` ship curated prompts and structured outputs validated by Pydantic.
+- Supports OpenAI and Azure OpenAI clients interchangeably, including async workloads and embeddings.
+
 ## 🚀 Quick Start: From Text to Insights in Seconds
 
 Imagine analyzing 10,000 customer reviews. Instead of manual work, just write:
@@ -85,21 +106,28 @@ This approach helps reduce latency and simplifies your code.
 Additionally, it integrates effortlessly with Pandas DataFrames and Apache Spark UDFs, making it easy to incorporate
 into your data processing pipelines.
 
-## Features
+Behind the scenes, `BatchingMapProxy` and `AsyncBatchingMapProxy` deduplicate repeated inputs, guarantee response order,
+and unblock waiters even when upstream APIs error. Caches created via helpers such as `responses_with_cache` plug into
+this batching layer so expensive prompts are reused across pandas, Spark, and async flows. Progress bars surface
+automatically in notebook environments when `show_progress=True`.
+
+## Core Capabilities
 
-- Vectorized API requests for processing multiple inputs at once.
-- Seamless integration with Pandas DataFrames.
-- A UDF builder for Apache Spark.
-- Compatibility with multiple OpenAI clients, including Azure OpenAI.
+- Vectorized request batching with automatic deduplication, retries, and cache hooks for any OpenAI-compatible client.
+- pandas `.ai` and `.aio` accessors for synchronous and asynchronous DataFrame pipelines, including `ai.extract` helpers.
+- Task library with Pydantic-backed schemas for consistent structured outputs across pandas and Spark jobs.
+- Spark UDF builders (`responses_udf`, `embeddings_udf`, `parse_udf`, `task_udf`, etc.) for large-scale ETL and BI.
+- Embeddings, token counting, and similarity utilities for search and retrieval use cases.
+- Prompt tooling (`FewShotPromptBuilder`, `improve`) to craft and iterate production-ready instructions.
 
 ## Key Benefits
 
-- **🚀 Performance**: Vectorized processing handles thousands of records in minutes, not hours
-- **💰 Cost Efficiency**: Automatic deduplication significantly reduces API costs on typical datasets
-- **🔗 Integration**: Works within existing pandas/Spark workflows without architectural changes
-- **📈 Scalability**: Same API scales from exploratory analysis (100s of records) to production systems (millions of records)
-- **🎯 Pre-configured Tasks**: Ready-to-use task library with optimized prompts for common use cases
-- **🏢 Enterprise Ready**: Microsoft Fabric integration, Apache Spark UDFs, Azure OpenAI compatibility
+- **🚀 Throughput**: Smart batching and concurrency tuning process thousands of records in minutes, not hours.
+- **💰 Cost Efficiency**: Input deduplication and optional caches cut redundant token usage on real-world datasets.
+- **🛡️ Reliability**: Guardrails for reasoning models, informative errors, and automatic waiter release keep pipelines healthy.
+- **🔗 Integration**: pandas, Spark, async, and Fabric workflows share the same API surface—no bespoke adapters required.
+- **🎯 Consistency**: Pre-configured tasks and extractors deliver structured outputs validated with Pydantic models.
+- **🏢 Enterprise Ready**: Azure OpenAI parity, Microsoft Fabric walkthroughs, and Spark UDFs shorten the path to production.
 
 ## Requirements
 
@@ -119,7 +147,7 @@ If you want to uninstall the package, you can do so with:
 pip uninstall openaivec
 ```
 
-## Basic Usage
+## Core Workflows
 
 ### Direct API Usage
 

{openaivec-0.15.1 → openaivec-0.99.0}/src/openaivec/__init__.py

@@ -1,13 +1,18 @@
 from ._embeddings import AsyncBatchEmbeddings, BatchEmbeddings
 from ._model import PreparedTask
-from ._prompt import FewShotPromptBuilder
+from ._prompt import FewShotPrompt, FewShotPromptBuilder
 from ._responses import AsyncBatchResponses, BatchResponses
+from ._schema import InferredSchema, SchemaInferenceInput, SchemaInferer
 
 __all__ = [
     "AsyncBatchEmbeddings",
     "AsyncBatchResponses",
     "BatchEmbeddings",
     "BatchResponses",
+    "FewShotPrompt",
     "FewShotPromptBuilder",
+    "InferredSchema",
     "PreparedTask",
+    "SchemaInferenceInput",
+    "SchemaInferer",
 ]

{openaivec-0.15.1 → openaivec-0.99.0}/src/openaivec/_proxy.py

@@ -199,7 +199,7 @@ class BatchingMapProxy(ProxyBase[S, T], Generic[S, T]):
     # - If positive integer: Fixed batch size
     # - If <= 0: Process all items at once
     batch_size: int | None = None
-    show_progress: bool = False
+    show_progress: bool = True
     suggester: BatchSizeSuggester = field(default_factory=BatchSizeSuggester, repr=False)
 
     # internals
@@ -509,7 +509,7 @@ class AsyncBatchingMapProxy(ProxyBase[S, T], Generic[S, T]):
     # - If <= 0: Process all items at once
     batch_size: int | None = None
     max_concurrency: int = 8
-    show_progress: bool = False
+    show_progress: bool = True
     suggester: BatchSizeSuggester = field(default_factory=BatchSizeSuggester, repr=False)
 
     # internals

{openaivec-0.15.1 → openaivec-0.99.0}/src/openaivec/pandas_ext.py

@@ -249,7 +249,7 @@ class OpenAIVecSeriesAccessor:
         instructions: str,
         response_format: type[ResponseFormat] = str,
         batch_size: int | None = None,
-        show_progress: bool = False,
+        show_progress: bool = True,
         **api_kwargs,
     ) -> pd.Series:
         """Call an LLM once for every Series element.
@@ -282,7 +282,7 @@ class OpenAIVecSeriesAccessor:
             batch_size (int | None, optional): Number of prompts grouped into a single
                 request. Defaults to ``None`` (automatic batch size optimization
                 based on execution time). Set to a positive integer for fixed batch size.
-            show_progress (bool, optional): Show progress bar in Jupyter notebooks. Defaults to ``False``.
+            show_progress (bool, optional): Show progress bar in Jupyter notebooks. Defaults to ``True``.
             **api_kwargs: Additional OpenAI API parameters (temperature, top_p, etc.).
 
         Returns:
@@ -341,7 +341,7 @@ class OpenAIVecSeriesAccessor:
             name=self._obj.name,
         )
 
-    def embeddings(self, batch_size: int | None = None, show_progress: bool = False, **api_kwargs) -> pd.Series:
+    def embeddings(self, batch_size: int | None = None, show_progress: bool = True, **api_kwargs) -> pd.Series:
         """Compute OpenAI embeddings for every Series element.
 
         Example:
@@ -362,7 +362,7 @@ class OpenAIVecSeriesAccessor:
             batch_size (int | None, optional): Number of inputs grouped into a
                 single request. Defaults to ``None`` (automatic batch size optimization
                 based on execution time). Set to a positive integer for fixed batch size.
-            show_progress (bool, optional): Show progress bar in Jupyter notebooks. Defaults to ``False``.
+            show_progress (bool, optional): Show progress bar in Jupyter notebooks. Defaults to ``True``.
             **api_kwargs: Additional OpenAI API parameters (e.g., dimensions for text-embedding-3 models).
 
         Returns:
@@ -417,7 +417,7 @@ class OpenAIVecSeriesAccessor:
         self,
         task: PreparedTask,
         batch_size: int | None = None,
-        show_progress: bool = False,
+        show_progress: bool = True,
     ) -> pd.Series:
         """Execute a prepared task on every Series element.
 
@@ -447,7 +447,7 @@ class OpenAIVecSeriesAccessor:
             batch_size (int | None, optional): Number of prompts grouped into a single
                 request to optimize API usage. Defaults to ``None`` (automatic batch size
                 optimization based on execution time). Set to a positive integer for fixed batch size.
-            show_progress (bool, optional): Show progress bar in Jupyter notebooks. Defaults to ``False``.
+            show_progress (bool, optional): Show progress bar in Jupyter notebooks. Defaults to ``True``.
 
         Note:
             The task's stored API parameters are used. Core batching / routing keys
@@ -520,7 +520,7 @@ class OpenAIVecSeriesAccessor:
         response_format: type[ResponseFormat] | None = None,
         max_examples: int = 100,
         batch_size: int | None = None,
-        show_progress: bool = False,
+        show_progress: bool = True,
         **api_kwargs,
     ) -> pd.Series:
         """Parse Series values into structured data using an LLM.
@@ -544,7 +544,7 @@ class OpenAIVecSeriesAccessor:
             batch_size (int | None, optional): Number of requests to process
                 per batch. None enables automatic optimization. Defaults to None.
             show_progress (bool, optional): Display progress bar in Jupyter
-                notebooks. Defaults to False.
+                notebooks. Defaults to True.
             **api_kwargs: Additional OpenAI API parameters (temperature, top_p,
                 frequency_penalty, presence_penalty, seed, etc.).
 
@@ -770,7 +770,7 @@ class OpenAIVecDataFrameAccessor:
         instructions: str,
         response_format: type[ResponseFormat] = str,
         batch_size: int | None = None,
-        show_progress: bool = False,
+        show_progress: bool = True,
         **api_kwargs,
     ) -> pd.Series:
         """Generate a response for each row after serializing it to JSON.
@@ -801,7 +801,7 @@ class OpenAIVecDataFrameAccessor:
             batch_size (int | None, optional): Number of requests sent in one batch.
                 Defaults to ``None`` (automatic batch size optimization
                 based on execution time). Set to a positive integer for fixed batch size.
-            show_progress (bool, optional): Show progress bar in Jupyter notebooks. Defaults to ``False``.
+            show_progress (bool, optional): Show progress bar in Jupyter notebooks. Defaults to ``True``.
             **api_kwargs: Additional OpenAI API parameters (temperature, top_p, etc.).
 
         Returns:
@@ -840,7 +840,7 @@ class OpenAIVecDataFrameAccessor:
         self,
         task: PreparedTask,
         batch_size: int | None = None,
-        show_progress: bool = False,
+        show_progress: bool = True,
     ) -> pd.Series:
         """Execute a prepared task on each DataFrame row after serializing it to JSON.
 
@@ -874,7 +874,7 @@ class OpenAIVecDataFrameAccessor:
             batch_size (int | None, optional): Number of requests sent in one batch
                 to optimize API usage. Defaults to ``None`` (automatic batch size
                 optimization based on execution time). Set to a positive integer for fixed batch size.
-            show_progress (bool, optional): Show progress bar in Jupyter notebooks. Defaults to ``False``.
+            show_progress (bool, optional): Show progress bar in Jupyter notebooks. Defaults to ``True``.
 
         Note:
             The task's stored API parameters are used. Core batching / routing
@@ -939,7 +939,7 @@ class OpenAIVecDataFrameAccessor:
         response_format: type[ResponseFormat] | None = None,
         max_examples: int = 100,
         batch_size: int | None = None,
-        show_progress: bool = False,
+        show_progress: bool = True,
         **api_kwargs,
     ) -> pd.Series:
         """Parse DataFrame rows into structured data using an LLM.
@@ -960,7 +960,7 @@ class OpenAIVecDataFrameAccessor:
             batch_size (int | None, optional): Rows per API batch. None
                 enables automatic optimization. Defaults to None.
             show_progress (bool, optional): Show progress bar in Jupyter
-                notebooks. Defaults to False.
+                notebooks. Defaults to True.
             **api_kwargs: Additional OpenAI API parameters.
 
         Returns:
@@ -1084,7 +1084,7 @@ class OpenAIVecDataFrameAccessor:
         target_column_name: str,
         max_examples: int = 500,
         batch_size: int | None = None,
-        show_progress: bool = False,
+        show_progress: bool = True,
     ) -> pd.DataFrame:
         """Fill missing values in a DataFrame column using AI-powered inference.
 
@@ -1102,7 +1102,7 @@ class OpenAIVecDataFrameAccessor:
             batch_size (int | None, optional): Number of requests sent in one batch
                 to optimize API usage. Defaults to ``None`` (automatic batch size
                 optimization based on execution time). Set to a positive integer for fixed batch size.
-            show_progress (bool, optional): Show progress bar in Jupyter notebooks. Defaults to ``False``.
+            show_progress (bool, optional): Show progress bar in Jupyter notebooks. Defaults to ``True``.
 
         Returns:
             pandas.DataFrame: A new DataFrame with missing values filled in the target
@@ -1251,7 +1251,7 @@ class AsyncOpenAIVecSeriesAccessor:
         response_format: type[ResponseFormat] = str,
         batch_size: int | None = None,
         max_concurrency: int = 8,
-        show_progress: bool = False,
+        show_progress: bool = True,
         **api_kwargs,
     ) -> pd.Series:
         """Call an LLM once for every Series element (asynchronously).
@@ -1281,7 +1281,7 @@ class AsyncOpenAIVecSeriesAccessor:
                 based on execution time). Set to a positive integer for fixed batch size.
             max_concurrency (int, optional): Maximum number of concurrent
                 requests. Defaults to ``8``.
-            show_progress (bool, optional): Show progress bar in Jupyter notebooks. Defaults to ``False``.
+            show_progress (bool, optional): Show progress bar in Jupyter notebooks. Defaults to ``True``.
             **api_kwargs: Additional keyword arguments forwarded verbatim to
                 ``AsyncOpenAI.responses.parse`` (e.g. ``temperature``, ``top_p``,
                 ``max_output_tokens``, penalties, future parameters). Core batching keys
@@ -1360,7 +1360,7 @@ class AsyncOpenAIVecSeriesAccessor:
         )
 
     async def embeddings(
-        self, batch_size: int | None = None, max_concurrency: int = 8, show_progress: bool = False, **api_kwargs
+        self, batch_size: int | None = None, max_concurrency: int = 8, show_progress: bool = True, **api_kwargs
     ) -> pd.Series:
         """Compute OpenAI embeddings for every Series element (asynchronously).
 
@@ -1385,7 +1385,7 @@ class AsyncOpenAIVecSeriesAccessor:
                 based on execution time). Set to a positive integer for fixed batch size.
             max_concurrency (int, optional): Maximum number of concurrent
                 requests. Defaults to ``8``.
-            show_progress (bool, optional): Show progress bar in Jupyter notebooks. Defaults to ``False``.
+            show_progress (bool, optional): Show progress bar in Jupyter notebooks. Defaults to ``True``.
             **api_kwargs: Additional OpenAI API parameters (e.g., dimensions for text-embedding-3 models).
 
         Returns:
@@ -1467,7 +1467,7 @@ class AsyncOpenAIVecSeriesAccessor:
         task: PreparedTask,
         batch_size: int | None = None,
         max_concurrency: int = 8,
-        show_progress: bool = False,
+        show_progress: bool = True,
     ) -> pd.Series:
         """Execute a prepared task on every Series element (asynchronously).
 
@@ -1500,7 +1500,7 @@ class AsyncOpenAIVecSeriesAccessor:
                 optimization based on execution time). Set to a positive integer for fixed batch size.
             max_concurrency (int, optional): Maximum number of concurrent
                 requests. Defaults to 8.
-            show_progress (bool, optional): Show progress bar in Jupyter notebooks. Defaults to ``False``.
+            show_progress (bool, optional): Show progress bar in Jupyter notebooks. Defaults to ``True``.
 
         Note:
             The task's stored API parameters are used. Core batching / routing
@@ -1575,7 +1575,7 @@ class AsyncOpenAIVecSeriesAccessor:
         max_examples: int = 100,
         batch_size: int | None = None,
         max_concurrency: int = 8,
-        show_progress: bool = False,
+        show_progress: bool = True,
         **api_kwargs,
     ) -> pd.Series:
         """Parse Series values into structured data using an LLM (asynchronously).
@@ -1594,7 +1594,7 @@ class AsyncOpenAIVecSeriesAccessor:
                 automatic optimization. Defaults to None.
             max_concurrency (int, optional): Maximum concurrent API requests.
                 Defaults to 8.
-            show_progress (bool, optional): Show progress bar. Defaults to False.
+            show_progress (bool, optional): Show progress bar. Defaults to True.
             **api_kwargs: Additional OpenAI API parameters.
 
         Returns:
@@ -1696,7 +1696,7 @@ class AsyncOpenAIVecDataFrameAccessor:
         response_format: type[ResponseFormat] = str,
         batch_size: int | None = None,
         max_concurrency: int = 8,
-        show_progress: bool = False,
+        show_progress: bool = True,
         **api_kwargs,
     ) -> pd.Series:
         """Generate a response for each row after serializing it to JSON (asynchronously).
@@ -1731,7 +1731,7 @@ class AsyncOpenAIVecDataFrameAccessor:
             **api_kwargs: Additional OpenAI API parameters (temperature, top_p, etc.).
             max_concurrency (int, optional): Maximum number of concurrent
                 requests. Defaults to ``8``.
-            show_progress (bool, optional): Show progress bar in Jupyter notebooks. Defaults to ``False``.
+            show_progress (bool, optional): Show progress bar in Jupyter notebooks. Defaults to ``True``.
 
         Returns:
             pandas.Series: Responses aligned with the DataFrame's original index.
@@ -1780,7 +1780,7 @@ class AsyncOpenAIVecDataFrameAccessor:
         task: PreparedTask,
         batch_size: int | None = None,
         max_concurrency: int = 8,
-        show_progress: bool = False,
+        show_progress: bool = True,
     ) -> pd.Series:
         """Execute a prepared task on each DataFrame row after serializing it to JSON (asynchronously).
 
@@ -1817,7 +1817,7 @@ class AsyncOpenAIVecDataFrameAccessor:
                 optimization based on execution time). Set to a positive integer for fixed batch size.
             max_concurrency (int, optional): Maximum number of concurrent
                 requests. Defaults to 8.
-            show_progress (bool, optional): Show progress bar in Jupyter notebooks. Defaults to ``False``.
+            show_progress (bool, optional): Show progress bar in Jupyter notebooks. Defaults to ``True``.
 
         Note:
             The task's stored API parameters are used. Core batching / routing
@@ -1884,7 +1884,7 @@ class AsyncOpenAIVecDataFrameAccessor:
         max_examples: int = 100,
         batch_size: int | None = None,
         max_concurrency: int = 8,
-        show_progress: bool = False,
+        show_progress: bool = True,
         **api_kwargs,
     ) -> pd.Series:
         """Parse DataFrame rows into structured data using an LLM (asynchronously).
@@ -1903,7 +1903,7 @@ class AsyncOpenAIVecDataFrameAccessor:
                 automatic optimization. Defaults to None.
             max_concurrency (int, optional): Maximum concurrent requests.
                 Defaults to 8.
-            show_progress (bool, optional): Show progress bar. Defaults to False.
+            show_progress (bool, optional): Show progress bar. Defaults to True.
             **api_kwargs: Additional OpenAI API parameters.
 
         Returns:
@@ -2036,7 +2036,7 @@ class AsyncOpenAIVecDataFrameAccessor:
         max_examples: int = 500,
         batch_size: int | None = None,
         max_concurrency: int = 8,
-        show_progress: bool = False,
+        show_progress: bool = True,
     ) -> pd.DataFrame:
         """Fill missing values in a DataFrame column using AI-powered inference (asynchronously).
 
@@ -2056,7 +2056,7 @@ class AsyncOpenAIVecDataFrameAccessor:
                 optimization based on execution time). Set to a positive integer for fixed batch size.
             max_concurrency (int, optional): Maximum number of concurrent
                 requests. Defaults to 8.
-            show_progress (bool, optional): Show progress bar in Jupyter notebooks. Defaults to ``False``.
+            show_progress (bool, optional): Show progress bar in Jupyter notebooks. Defaults to ``True``.
 
         Returns:
             pandas.DataFrame: A new DataFrame with missing values filled in the target

{openaivec-0.15.1 → openaivec-0.99.0}/src/openaivec/spark.py

@@ -150,6 +150,8 @@ from openaivec._serialize import deserialize_base_model, serialize_base_model
 from openaivec._util import TextChunker
 
 __all__ = [
+    "setup",
+    "setup_azure",
     "responses_udf",
     "task_udf",
     "embeddings_udf",
@@ -242,50 +244,6 @@ def setup_azure(
     CONTAINER.clear_singletons()
 
 
-def set_responses_model(model_name: str):
-    """Set the default model name for response generation in the DI container.
-
-    Args:
-        model_name (str): The model name to set as default for responses.
-    """
-    CONTAINER.register(ResponsesModelName, lambda: ResponsesModelName(model_name))
-    CONTAINER.clear_singletons()
-
-
-def get_responses_model() -> str | None:
-    """Get the default model name for response generation from the DI container.
-
-    Returns:
-        str | None: The default model name for responses, or None if not set.
-    """
-    try:
-        return CONTAINER.resolve(ResponsesModelName).value
-    except Exception:
-        return None
-
-
-def set_embeddings_model(model_name: str):
-    """Set the default model name for embeddings in the DI container.
-
-    Args:
-        model_name (str): The model name to set as default for embeddings.
-    """
-    CONTAINER.register(EmbeddingsModelName, lambda: EmbeddingsModelName(model_name))
-    CONTAINER.clear_singletons()
-
-
-def get_embeddings_model() -> str | None:
-    """Get the default model name for embeddings from the DI container.
-
-    Returns:
-        str | None: The default model name for embeddings, or None if not set.
-    """
-    try:
-        return CONTAINER.resolve(EmbeddingsModelName).value
-    except Exception:
-        return None
-
-
 def _python_type_to_spark(python_type):
     origin = get_origin(python_type)
 

{openaivec-0.15.1 → openaivec-0.99.0}/tests/test_proxy.py

@@ -608,7 +608,7 @@ def test_progress_bar_methods():
 
 
 def test_batching_proxy_with_progress_disabled():
-    """Test BatchingMapProxy with progress disabled (default)."""
+    """Test BatchingMapProxy with progress disabled via explicit flag."""
     calls: list[list[int]] = []
 
     def mf(xs: list[int]) -> list[int]:
@@ -641,7 +641,7 @@ def test_batching_proxy_with_progress_enabled():
 
 @pytest.mark.asyncio
 async def test_async_batching_proxy_with_progress_disabled():
-    """Test AsyncBatchingMapProxy with progress disabled (default)."""
+    """Test AsyncBatchingMapProxy with progress disabled via explicit flag."""
     calls: list[list[int]] = []
 
     async def mf(xs: list[int]) -> list[int]: